Wttr Weather

Wttr Weather is an MCP server that connects LLM interfaces and AI agents to the wttr.in weather service. Designed for developers, automation builders, and users operating AI assistants locally or remotely, it enables language models to fetch real-time atmospheric conditions and multi-day meteorological forecasts for specific cities worldwide. By exposing weather retrieval endpoints through the Model Context Protocol, the server bridges the gap between static model weights and dynamic environmental data without requiring complex third-party API keys. Users interacting with supported MCP clients can request current conditions, including temperature and weather states, or retrieve extended multi-day outlooks directly inside their prompt sessions. The server runs containerized via Docker to standardize runtime dependencies and isolation, drawing architectural patterns from community implementations like the DuckDuckGo MCP server. It has been validated using local Ollama model configurations such as Llama 3.2 and Qwen3, making it practical for private, local AI workflows as well as general assistant environments. Through straightforward city name queries, the integration supplies formatted meteorological context directly to conversational agents for travel planning, scheduling, or routine status reporting.

Category: Maps, Weather & Local Data

Tags: forecast, meteorology, weather, wttr

Visit Wttr Weather

How to install and configure Wttr Weather

1. Build the Docker container image locally by running the build script from the repository: bash ./build_docker_image.sh 2. Open your MCP client configuration file (for example, Claude Desktop's configuration file). 3. Register the server under the mcpServers section using the following snippet: json { "mcpServers": { "web_fetch_wttr": { "command": "docker", "args": [ "run", "--rm", "-i", "--init", "web_fetch_wttr:1.0.0" ] } } } 4. Save your configuration and restart your client.

What you can do with Wttr Weather

• Check current meteorological conditions for any named city directly within an ongoing assistant conversation. - Retrieve a three-day weather forecast to assist with travel planning, scheduling, or packing advice. - Provide real-time local weather context to local language models running via Ollama environments. - Automate daily summary prompts that combine current weather, multi-day projections, and routine scheduling data.

Key facts

• https://github.com/melody26613/mcp-server-fetch-wttr
• Maps, Weather & Local Data
• forecast, meteorology, weather, wttr

Part of MCP Servers

What can Wttr Weather do?

Wttr Weather allows MCP-compatible AI models to retrieve weather information from the wttr.in service. It provides tools to fetch the current weather conditions for any specified city as well as an extended three-day forecast, delivering live meteorological context straight into agent conversations.

How do I install Wttr Weather?

Wttr Weather is installed by building its Docker container using the provided shell script build_docker_image.sh. Once built, you add the server configuration specifying the docker run command and the web_fetch_wttr:1.0.0 image to your client settings file, then restart the client application.

What tools are included in Wttr Weather?

The server exposes two main tools: get_current_weather, which accepts a city name parameter to look up present conditions, and get_three_day_weather, which takes a city name to return a three-day forecast. Both return formatted text strings directly from the wttr.in service.

Which models or clients work with Wttr Weather?

Wttr Weather works with standard Model Context Protocol clients including Claude Desktop. The project documentation notes specific testing with local Ollama models, including llama3.2:3b-instruct-q2_K, qwen3:0.6b, and qwen3:1.7b, confirming compatibility across both proprietary client applications and lightweight local inference environments.